The 2005 IOM report on Complementary and Alternative Medicine notes that although CAM therapies have existed throughout history and are used by at least one-third of all Americans, "it is only relatively recently...that there has been a serious general interest...in investigating and evaluating these therapies." Indeed, even the epidemiology of the most recognizable form of CAM "chiropractic" is not well described, although its use by older adults has been increasing since 2000. Accordingly, in this revised R21 application (submitted in response to PAR 08-213, Methodology and Measurement in the Behavioral and Social Sciences), we propose to examine who uses chiropractic services, the volume and consistency of the chiropractic services that they use over time, and the antecedents and consequences of that chiropractic use among a nationally representative sample of Medicare beneficiaries. To do this, we have successfully linked data for 5,511 self-respondents who participated in the Study of Assets and Health Dynamics Among the Oldest Old (AHEAD) who were not in managed care Medicare at baseline (74% of the original cohort of 7,447 non-institutionalized subjects who were 70 years old or older at baseline). These 5,511 subjects are censored post-baseline at entrance into managed care Medicare or death, whichever comes first. The linked data sources are: (a) their 1993 baseline and 1995, 1998, 2000, 2002, 2004, and 2006 follow-up AHEAD interviews (b) geocode identifiers for each subject at each wave of data;and, (c) Medicare Part A and B claims for calendar years 1991-2005. With these data, we will address three specific aims: (1) to characterize chiropractic use patterns over time among older adults, (2) to identify antecedents of chiropractic use patterns;and, (3) to evaluate the consequences of chiropractic use patterns on physician visits, all-cause hospitalizations, hospitalization for ambulatory care sensitive conditions, Medicare payments, and mortality. Our methods will include intuitive classification rules and algorithms as well as latent class analyses (Aim 1), multivariable binomial and multinomial logistic regression analyses with pattern mixture models (Aim 2), and multivariable proportional hazards, binomial, and Poisson regression analyses with non-equivalent dependent variables (Aim 3). We are especially interested in (a) evaluating how the 2000 CMS reimbursement policy changes affected demand and use patterns (i.e., enduring effects on annual use rates and use patterns over time);(b) exploring the potential reimbursement gaming that we have identified to circumvent the arbitrary CMS cap of 12 visits during any 12-month period;(c) determining whether substitution of chiropractic for primary care occurs;and, (d) identifying patterns of use over time that can then be used to focus policy interventions designed to either shrink the proportion of "chronic high users" or establish objective guidelines for their approval. PUBLIC HEALTH RELEVANCE: As the 2005 IOM report noted, the use of complementary and alternative medicine (CAM) in the United States, including its most readily identifiable form-chiropractic-is widespread but not well understood. Indeed, prior to this application, there have been no studies of chiropractic use patterns over time, or their antecedents and consequences among older Americans. In this R21 application, therefore, we propose to do exactly what the 2005 IOM report recommended-to conduct an extensive, secondary analysis of the nationally representative sample of older adults in the AHEAD, which is a large, prospective cohort study.